Differences in Temperature Measurements by Commercial Temperature Sensors

The accuracy of room temperature measurements is critical for the control of indoor thermal environment as the HVAC system uses this as the input and control target. The temperature measurements can vary largely depending on several factors. The present study summarizes the temperature measurement performance of several commercially available wired, and wireless temperature sensors.

Keywords: Building Management System (BMS), Indoor Environmental Quality (IEQ), Wired and wireless temperature sensor, Control.

Introduction

Indoor thermal environments in buildings are typically controlled by Building Management Systems (BMS) that use temperature readings either from sensors placed on walls in rooms or in exhaust ducts. The readings from these temperature sensors are critical as the HVAC system is controlled based on this input (sometimes together with other inputs such as relative humidity and CO₂ concentrations). However, these temperature readings have limitations due to their location and measurement capabilities.

One recent trend is the possibility of replacing wired, wall-mounted sensors with wireless sensors, that could typically measure temperature, relative humidity and CO₂ concentration, by deploying them in large numbers and close to occupants (i.e., in the occupied zone instead of on walls or in exhaust ducts). However, these sensors vary widely in terms of their shape, casing, interior and exterior materials, colour, size, openings and the internal placement of different sensing components. All these factors influence the temperature reading. Although wired, wall-mounted room temperature sensors tend to be more “uniform” in terms of their colour, size, shape, and openings, they are placed on walls (not in the occupied zone) and their placement on walls is typically dictated by the location of existing or available electrical installations.

Due to these characteristics of wired and wireless indoor environment sensors, it is crucial to identify and quantify their capabilities in accurately measuring the thermal environments (here simplified to temperature) in occupied zones, which they are expected to represent. The present article is based on two previous scientific journal articles by Mylonas et al. [1] and Shinoda et al. [2]. These two articles are briefly introduced, and their results are summarized with a focus on the performance of wired and wireless indoor environment sensors in terms of temperature measurement.

Measurements with wireless sensors

To identify whether wireless indoor environment sensors could reliably replace wired sensors, Mylonas et al. [1] carried out a measurement campaign in a climate chamber. The main goal of their study was to investigate the performance of different wireless indoor environment sensors in terms of their accuracy in measuring CO₂ concentration, temperature and relative humidity, and the factors affecting their CO₂ measurement error. The measurements of temperature, relative humidity and CO₂ concentration were compared to the measurements obtained with lab-grade, high precision reference instruments.

Mylonas et al. [1] identified six sensor types to be tested out of forty-four commercially available wireless indoor environment sensors at the time. The selection criteria included temperature, relative humidity and CO₂ concentration measurement capability, internet connectivity, real-time remote tracking possibility, ease of access to measured data, availability and price. Three identical sensors from each of the selected six sensor types were tested.

The experiments were carried out in a climate chamber at the International Centre for Indoor Environment and Energy, Technical University of Denmark. The sensors were placed in the climate chamber on a table either horizontally or vertically according to the manufacturer recommendations.

Figure 1 shows the experimental setup.

Figure 1. Experimental setup with wireless sensors (adopted from [1]).

The experiments were carried out under steady-state conditions. Temperature and CO₂ concentrations were controlled. The humidity in the chamber was measured but not controlled. Test conditions were chamber temperatures of 16, 20, 25 and 30°C, and CO₂ concentrations of 400, 800, 1 200, 1 600 and 1 930 ppm. Different calibration temperatures of 20 and 30°C were also tested for the CO₂ sensors.

The authors compared the measurements obtained by the wireless sensors to the measurements obtained by the reference instruments and compared these results to relevant standards and manufacturer specified measurement accuracies.

Measurements with wireless, and wired sensors

Building on the study by Mylonas et al. [1], Shinoda et al. [2] studied the temperature measuring performance of nine wireless and two conventional wired temperature sensors by comparing their measurements to reference air and globe temperatures in a climate chamber with a two-person office setup at the Technical University of Denmark.

Shinoda et al. [2] studied the same wireless sensors that were evaluated by Mylonas et al. [1], together with the four wall-mounted (two wireless and two wired) sensors that were evaluated by Borier et al. [3]. The four wall-mounted sensors were selected to represent conventional and new generation thermostats. Measured values were compared to the reference air and globe temperatures.

Figure 2 shows the sensors used in the experiments of Shinoda et al. [2].

Figure 2. The external casing of the sensors used in the experiments (adopted from [2]).

Most of the sensors had a white casing. Sensors D1 and D2 had a silver metallic casing, and sensor W had a silver metallic casing on the side and a black screen on the front. Some sensor casings did not have any openings for air to pass through, while others (such as sensors E and F) had a wide opening (e.g. a grill).

Shinoda et al. [2] studied the influence of sensor type and position (location in the chamber and height above the floor), room cooling system (all-air cooling system, and radiant ceiling panels combined with mixing ventilation) and cooling load (33 and 61 W/m²) on the temperature measurements. In addition to the typical internal heat gains (i.e., occupants, equipment and lighting), heat gains from a heated surface (window) due to solar radiation was emulated and there were electric mats simulating heated floor close to the window (to emulate the direct solar radiation coming through the window, hitting the floor surface, and heating the floor surface).

Summary of the findings

Obtaining accurate readings from temperature sensors in indoor spaces is critical for the optimal control of HVAC systems to obtain the optimal indoor environment for the occupants with minimal energy use.

The tested wireless sensors performed satisfactorily in terms of temperature measurement i.e., they met the measurement criteria according to ISO 7726 [4] and manufacturer specifications. This indicates that the studied wireless indoor environment sensors can potentially be used to control the temperature in buildings; however, this does not apply to the relative humidity and CO₂ concentration measurements. Significant improvements in relative humidity and CO₂ concentration measurement capabilities are required before these sensors can be installed in buildings for humidity and CO₂ concentration control.

The measurements in the realistically setup climate chamber showed that sensors placed at the same position had a measurement difference of up to 1.8 K. This is a large difference considering that Category II of EN 16798-2 [5] suggests a temperature range of 23 – 26°C for cooling operation. A difference of 1.8 K within this range can influence the control and classification of the thermal indoor environment significantly and shows the importance of accurate temperature sensing.

Different sensitivities to radiation, casing properties, and the internal distribution of different sensing components affected the temperatures measured by different sensors. The assumption about the type of temperature (i.e., air or globe) a sensor measures had the largest impact on the deviation from the reference temperature. Wireless sensors varied in their sensitivity to radiation and differed in their relationship to the reference air and globe temperature measurements. In most cases, the temperature difference from the reference air or globe temperature was within ±0.5 K. If the difference between the mean radiant temperature and the air temperature is negligible, any of the studied sensors could be used as a temperature sensing component of the BMS, but when a large difference between the air and mean radiant temperature is expected at a given position (e.g., in perimeter zones), it is suggested that sensors that measure closer to globe temperature are used.

Under the high cooling load conditions, measurements in radiant system scenarios had smaller deviations from the reference sensors compared to all-air system scenarios, due to the chilled ceiling surface compensating for the radiation from the room loads. This finding agrees with the previous findings of Kazanci et al. [6].

As opposed to typical assumptions, conventional wired temperature sensors measured closer to the globe temperature than to the air temperature, mainly due to their casing properties such as shape and material. Most measured values of the wired sensors were within their manufacturer-stated accuracy range if the temperatures measured are assumed to be the globe temperature rather than the air temperature. This indicates that in practice, measurements from a conventional wired temperature sensor may be assumed to represent the operative temperature (estimated as the globe temperature) if it is located at a height of 0.6 or 1.1 m. Measurements at a height of 1.5 m (which is more typical for wired BMS sensors), if it cannot be placed at 0.6 or 1.1 m, can also be used as a proxy for the operative temperature in most cases. In rooms equipped with a conventional wired thermostat, HVAC systems are thus controlled by measurements close to operative temperature. When performing simulations, the control reference should be set as the operative temperature instead of air temperature. However, the possibility of a difference in the air and mean radiant temperature between the thermostat position and the different locations in the occupied zone must still be considered.

It is recommended that manufacturers conduct measurement accuracy tests of their sensors in a climate chamber with a temperature difference between the air and mean radiant temperature to test their sensors’ response to radiation. In general, the testing conditions should cover temperature, relative humidity, and CO₂ concentration ranges indicated in the most prominent EN and ISO standards (e.g., EN 16798-1 [7] and ISO 17772-1 [8]) for indoor environmental quality. Testing under certain extreme conditions should also be considered depending on the intended use of the specific sensor. It is also recommended that tests are carried under transient conditions so that the response of sensors to dynamic indoor environmental conditions can be quantified.

Only the main findings from [1] and [2] are reported in the present article and the reader is referred to the original publications for further details. It should be noted that the two scientific articles [1,2], which the present article is based on, were published five and three years ago, respectively. It is likely that there have been advancements in both wired and wireless indoor environment sensor technology since then. Nevertheless, the findings of these articles remain valid and should be considered for future sensor selection, development, testing, and application.

Acknowledgements

This article is based on two previous publications referenced as [1] and [2]. The author thanks all co-authors of those two publications.

References

[1]     Mylonas, A., Kazanci, O. B., Andersen, R. K., & Olesen, B. W. (2019). Capabilities and limitations of wireless CO₂, temperature and relative humidity sensors. Building and Environment, 154, 362-374. doi: 10.1016/j.buildenv.2019.03.012.

[2]     Shinoda, J., Mylonas, A., Kazanci, O. B., Tanabe, S. I., & Olesen, B. W. (2021). Differences in temperature measurement by commercial room temperature sensors: Effects of room cooling system, loads, sensor type and position. Energy and Buildings, 231, 110630.doi: 10.1016/j.enbuild.2020.110630.

[3]     O.M. Borier, O.B. Kazanci, B.W. Olesen, D. Khovalyg, Which sensor type at which location should offices with south orientated window choose to improve comfort and reduce energy consumption?, J. Phys.: Conf. Ser. 1343 (2019) 012147.

[4]     EN ISO 7726: Ergonomics of the thermal environment - Instruments for measuring physical quantities, European Committee for Standardization, Brussels, 2001.

[5]     EN 16798-2: Energy performance of buildings - Ventilation for buildings - Part 2: Interpretation of the requirements in EN 16798-1 - Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics, European Committee for Standardization, Brussels, 2019.

[6]     O.B. Kazanci, D. Khovalyg, T. Iida, Y. Uno, T. Ukiana, B.W. Olesen, Experimental comparison of the thermal indoor environment created by a radiant, and a combined radiant and convective cooling system, in: Proceedings— Roomvent & Ventilation 2018, 2018: pp. 223–228.

[7]     EN 16798-1: Energy performance of buildings - Ventilation for buildings - Part 1: Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics, European Committee for Standardization, Brussels, 2019.

[8]     ISO 17772-1: Energy performance of buildings - Indoor environmental quality - Part 1: Indoor environmental input parameters for the design and assessment of energy performance of buildings, 2017.